Several variables, including age, are known to influence anxiety. Previous exposure to the elevated-plus maze (EPM) is known to modify emotional behaviour as retesting in the EPM at a standard age of 3 months increases open-arm avoidance and attenuates the effects of anxiolytic drugs. This study analysed whether similar results are obtained when older animals are subjected to these experimental paradigms. Overall, increasing age was associated with more signs of anxiety. Additionally, we observed a paradoxical behaviour pattern in aged-subjects that were re-exposed to the EPM, with mid-aged and old rats failing to display open arm avoidance (OAA) in the second trial; this qualitative shift in emotional behaviour was not associated with decreased locomotion. An examination of how age influences responsiveness to anxiolytic drugs, with or without previous maze experience, was also conducted. Midazolam (0.5 and 1 mg/kg) proved anxiolytic in maize-naive young animals; in marked contrast, in older animals midazolam at 1 mg/kg resulted in sedation but not anxiolyis. One trial tolerance to midazolam was evident in animals of both ages that were subjected to a second EPM trial; the latter phenomenon was apparently accentuated in older animals as they do not show open arm avoidance upon re-exposure to the EPM. These data suggest that the age-associated 'resistance' to anxiolytic drugs might be related to a qualitative shift in emotional behaviour.